3_HRcompare2

Fig. 3. Video of the comparison with RecycleGAN: we aim to compare the content consistency and image-level quality.

Here the RecycleGAN results are produced by the original RecycleGAN code in a resolution of 256×256. Since there is no guarantee of style consistency for RecycleGAN, we select some Re-cycleGAN visual results in a small sequence length of 30 frames where style is almost consistent to compare with UVIT (ours).

UVIT
UVIT
Unsupervised Multimodal Video-to-Video Translation via Self-Supervised Learning

Related